The Virgo detector. L. Rolland LAPP-Annecy GraSPA summer school L. Rolland GraSPA2013 Annecy le Vieux

Transcription

1 The Virgo detector The Virgo detector L. Rolland LAPP-Annecy GraSPA summer school

2 Table of contents Principles Effect of GW on free fall masses Basic detection principle overview Are the Virgo mirrors free fall masses? Virgo optical configuration, or how to measure m? Simple Michelson interferometer How do we improve the detector sensitivity? How do we measure the GW strain, h(t), from this detector? Some noises of the Virgo detector What is a noise? The fundamental noises: seismic, thermal, and shot noises History of Virgo noise 2

3 The Virgo detector Principles Reminder: effect of a GW on free masses A gravitational wave (GW) modifies the distance between free fall masses Case of a GW with polarization + propagating along z 3

4 The Virgo detector Principles A general overview of the Virgo detector 3 km arms! 4

5 The Virgo detector Principles Virgo: a more complicated interferometer Suspended mirrors Fabry-Perot cavities L L hl Ly = 2 Lx = + hl 2 Photodiodes Laser Infra-red laser P ~ 100 W Power-recycling cavity Sensor: photodiode we do not image the interference pattern! 5

6 The Virgo detector Principles Why are the Virgo mirrors free masses? We want the mirrors (mass M) to be free falling masses: In the case of sinusoidal regime: Mass M can be considered as free along x if 6

7 The Virgo detector Principles The case of the Virgo mirrors Mirrors can be considered as free for frequencies larger than ~10 Hz Input beam Transmitted beam 7

8 The Virgo detector Principles How and for what did you use interferometers? Wavelength of monochromatic source Sodium doublet wavelength separation Classroom interferometer Virgo interferometer Pisa, Italy 8

9 The Virgo detector Optical configuration Virgo optical configuration Reminder about planes waves How do we observe L with a Michelson interferometer? Measurement of a power variations From power variations to L (or to gravitational wave amplitude h) Improving the interferometer: How do we increase the power on the beam splitter mirror? How do we amplify the phase offset between the arms? 9

10 The Virgo detector Optical configuration Description of plane waves Plane wave propagating along z, with speed c amplitude wavelength (m) wave number (rad/m) angular frequency (rad/s) Average power: Complex form > simpler algebraic calculations, for example > real plane wave is the real part: Plane waves do not exist but they are a good approximation of many waves in localized region of space 10

11 The Virgo detector Optical configuration How do we observe L with a Michelson interferometer? Input wave BS located at (0,0) Sensor located at (0, ys) Input beam Amplitude reflection and transmission coefficients: and Beam splitter (BS) Transmitted beam Sensor y x We are interested in the beam transmitted by the interferometer: it is the sum of the two beams (fields) that have propagated along each arm. Around the beam splitter mirrors: Radius of curvature of the beams ~ 1400 m Size of the beams ~ few cm The beams can be approximated by plane waves 11

12 The Virgo detector Optical configuration How do we observe L with a Michelson interferometer? Input wave Beam propagating along x arm: Input beam Beam splitter (BS) Transmitted beam Sensor y x Sign convention for amplitude reflection and transmission coefficients 12

13 The Virgo detector Optical configuration How do we observe L with a Michelson interferometer? Input wave Beam propagating along x arm: Input beam Beam splitter (BS) Transmitted beam Sensor y x Sign convention for amplitude reflection and transmission coefficients 13

14 The Virgo detector Optical configuration How do we observe L with a Michelson interferometer? Input wave Beam propagating along x arm: Input beam Beam splitter (BS) Transmitted beam Sensor y x Sign convention for amplitude reflection and transmission coefficients 14

15 The Virgo detector Optical configuration How do we observe L with a Michelson interferometer? Input wave Beam propagating along x arm: Input beam Beam splitter (BS) Transmitted beam Sensor Complex reflection of the x arm y x Sign convention for amplitude reflection and transmission coefficients 15

16 The Virgo detector Optical configuration How do we observe L with a Michelson interferometer? Input wave Beam propagating along x arm: Beam splitter (BS) Input beam Transmitted beam Sensor Complex reflection of the x arm Beam propagating along y arm: y x + Transmitted field: Complex reflection of the y arm 16

17 The Virgo detector Optical configuration Power transmitted by a simple Michelson Transmitted field: Calculation of the transmitted power: With C=1 With C=0.5 17

18 The Virgo detector Optical configuration What power does Virgo measure? In general, the beam is not a plane wave but a spherical wave interference pattern (and the complementary pattern in reflection) Virgo interference pattern much larger than the beam size: ~1 m between 2 two consecutive fringes we do not study the fringes in nice images! Equivalent size of Virgo beam Setting a working point With C=1 Controlled mirror positions Freely swinging mirrors 18

19 The Virgo detector Optical configuration From the power to the gravitational wave Around the working point: Power variations as function of small differential length variations: 19

20 The Virgo detector Optical configuration From the power to the gravitational wave Around the working point: (W/m) Measurable physical quantity Physical effect to be detected 20

21 The Virgo detector Optical configuration Improving the interferometer sensitivity Increase the phase difference between the arms for a given differential arm length variation Increase the input power Fabry-Perot cavities in the arms Recycling cavity BS 21

22 The Virgo detector Optical configuration Optical cavity with two mirrors Cavity made of two plane infinite mirrors, in front of each other. z Sign convention for amplitude reflection and transmission coefficients 22

23 The Virgo detector Optical configuration Optical cavity with two mirrors Cavity made of two plane infinite mirrors, in front of each other. z Sign convention for amplitude reflection and transmission coefficients Relations between the fields at input and output of the cavity: 23

24 The Virgo detector Optical configuration In Virgo, the beam is resonant inside the cavities Cavity field as function of input field: z Power in the cavity: Virgo F = 50 AdVirgo F = 443 Airy peaks Virgo cavity at resonance: 24

25 The Virgo detector Optical configuration Field reflected by a Virgo arm cavity Reflected field as function of input field: Power reflected by the cavity, with Phase of the field reflected by one arm cavity around resonance: z Cavity around resonance Field reflected by the x arm: 25

26 The Virgo detector Optical configuration How do we amplify the phase offset? 3 km Fabry Perot cavities Input beam Input beam BS BS Transmitted beam Transmitted beam Sensor Sensor ~number of round trips in the arm ~300 for AdVirgo (instead of in the arm of a simple Michelson) 26

27 The Virgo detector Optical configuration How do we increase the power on BS? Detector working point close to a dark fringe most of power go back towards the laser Input beam BS Power recycling cavity Transmitted beam Resonant power recycling cavity input power on BS increased by a factor 38! 27

28 The Virgo detector Optical configuration The improved interferometer response Response of simple Michelson: (W/m) 3 km Fabry Perot cavities Input beam Response of recycled Michelson with Fabry Perot cavities: ~38 BS Power recycling cavity ~300 Transmitted beam Sensor 28

29 The Virgo detector Optical configuration A hint of AdvancedVirgo sensitivity Response of recycled Michelson with Fabry Perot cavities: 3 km Fabry Perot cavities Input beam BS Power recycling cavity Transmitted beam Sensor In reality, the detector response depends on frequency... 29

30 The Virgo detector Optical configuration Optical layout of Virgo 30

31 The Virgo detector How do we measure the GW strain, h(t), from this detector? How do we control the working point? We want to be (almost) fixed! Control loop done for noises with f between ~10 Hz and ~100 Hz Precision of the control ~ m Noises Input beam Transmitted beam Real-time digital calculations 31

32 The Virgo detector How do we measure the GW strain, h(t), from this detector? From the data to the GW strain h(t)... Transmitted power variations (W) Control signals (V) Input signals + (m/w) attenuated by controls as if no control, as if free falling mirrors (m/v) Responses to be measured (calibrated) in dedicated datasets 32

33 The Virgo detector How do we measure the GW strain, h(t), from this detector? AdVirgo data acquisition summary Photodiodes Currents in the mirror actuators Sismometers Magnetometers Thermometers Cameras Sensors... Synchronized digitization Sampling frequencies from 1 Hz to ~50 khz Data acquisition processes (data formatting, timestamp, write to disks..) ~ channels Continuous flow of ~2 TBytes/day (20 to 40 MBytes/s) Disk space on Virgo site: ~400 TB for 6 months of data Longer storage: data sent via Ethernet to computing centers (Lyon, Bologna) 33

34 Virgo noises 34

35 The Virgo detector Noises What is a noise in Virgo? Stochastic (random) signal that contributes to the signal h_mes(t) but does not contain information on the gravitational wave strain h_gw(t) Extracted from Black Hole Hunter: 35

36 How do we characterize a noise? Hypothesis: - we are looking for a constant signal S0 in the data - data are noisy (Gaussian noise) Projection of noise data Noise (units) Data points of noise only Noise (units) Noise (units) Time (s) Gaussian distribution: The mean value of the noise stays around 0 The mean value of the signal stays around S0. The variations of the noise decrease when the data are averaged over longer time Noise (units) Noise (units) Time (s) What is important to characterize a noise is it dispersion σnoise! Noise (units) 36

37 How do we characterize a noise? Projection of noise data Noise (units) Data points of noise only Noise (units) Noise (units) Time (s) The variations of the noise decrease when the data are averaged over longer time The noise can be characterized by the coefficient of proportionality D Noise (units) 37

38 How do we characterize a noise...in frequency domain? Fourier transform 38

39 The Virgo detector Noises What is the noise level of Virgo? h(f) [1/sqrt(Hz)] (a) Virgo Nominal sensitivity (b) Seismic noise (c) Pendulum thermal noise (d) Mirror thermal noise (e) Shot Noise (b) (d) (a) (e) 10 (c) Frequency [Hz] 39

40 The Virgo detector Noises Seismic noise and suspended mirrors Ground vibrations up to ~1 µm at low frequency decreasing down to ~ 10 pm at 100 Hz Transfer function 40

41 The Virgo detector Noises Seismic noise and the Virgo suspension Passive attenuation: 7 pendulum in cascade Active controls at low frequency 7m Accelerometers or interferometer data Electromagnetic actuators Control loops 41

42 The Virgo detector Noises Some noises: thermal noise Microscopic thermal fluctuations > dissipation of energy through excitation of the macroscopic modes of the mirror Pendulum mode f < 40 Hz Mirror modes 40 Hz < f < 100 Hz Violin modes We want high quality factors Q to concentrate all the noise in a small frequency band 42

43 The Virgo detector Noises What is the shot noise? Fluctuations of arrival times of photons (quantum noise) 43

44 The Virgo detector Noises Some other noises Acoustic vibrations and refraction index fluctuations Laser: amplitude, frequency, jitter noise Lots of control loops to reduce these noises Electronics noise Main elements installed in vacuum Challenge for the electronicians to measure down to 0.1 W/sqrt(Hz) Non linear noise from diffuse light Need dedicated optical elements with specific mechanical modes 44

45 The Virgo detector Noises History of Virgo noise curve 45

46 The Virgo detector Noises Noises are not always stationary... Glitches are impulses of noise. They might look like a transient GW signal... Now it is time to play with the data analysis! 46